Genetic variation associated with

chronic disease susceptibility in the

Portuguese population

Marta Barreto

Instituto Nacional de Saúde Doutor Ricardo Jorge

June 27th 2013

Genetic variation

Objectives

• Identify genetic factors that influence the

risk of prevalent chronic diseases in the

Portuguese population;

• Characterize the contribution of different

genetic factors to chronic disease

susceptibility.

4

1) Population sample

Materials and methods

•221 participants (95 men and 126 women).

•Pilot study of INSEF – “Inquérito Nacional de Saúde com exame Fisico” - the National Component of the European Health Examination Survey project. • São Brás de Alportel Health Center • Random sampling of participants by SNS identification number •In accordance with EHES procedures to achieve maximum participating rates and quality of the data and samples

http://www.google.pt/url?sa=i&rct=j&q=sampling&source=images&cd=&cad=rja&docid=CoB4BsVN0jFk-M&tbnid=T_HlQU0Hn395yM:&ved=0CAUQjRw&url=http://www.floodlightsurveys.com/blog/2011/06/sampling-stratified/&ei=3EWvUbXVGqaW0QX50YHoAQ&bvm=bv.47380653,d.ZGU&psig=AFQjCNGCHJR4Ez_LpE83LZSO1MMzO2VJ6g&ust=1370527568985819

Materials and methods 2) Phenotype characterization: Detailed Questionnaire, (sociodemographics

and occupation, medical history and general

health, family history of illness focusing on chronic

disorders, psychological status, and lifestyle

exposures (including smoking, alcohol, physical

activity and diet)

Physical exam (weight, height, waist and hip circumpherence, blood pressure)

Blood sample (Glucose, HDL, LDL, Triglycerides, GGT, ALT, AST, Creatinine, C

Reactive protein) CBC + Serum, Plasma and

DNA for Biobanking

3) Candidate gene analysis Blood sample DNA Extraction

SNP analysis

3) Candidate gene analysis

3.1 Candidate Genes SNPs selection (described in literature) 82 diferent Genes associated to: • Cancer • Drug resistance/metabolism • Cardiovascular diseases • Diabetes • Obesity • Psychiatric disorders • Drug addiction

RFLPs

Potential Public Health Impact

Sequenom-Massarrays

3.2 Genotyping:

Continuous Variables

HES Database

Genotypes

Associations

Lifestyle behaviours

Genotype Database $

To identify genetic risk factors involved in chronic disease susceptibility, using continuous variables.

Phenotype Database $

Materials and methods

Population characterization

Total Men Women P-value*

Number of participants 206 87 (42,2%) 119 (57,8%)

Age (years±SD) 56,31 ± 16,37 55,80 ± 16,45 56,67 ± 16,37 0,754

BMI(Kg/m2) 27,88 ± 4,69 27,44± 4,20 28,20± 5,01 0,336

MetS 1 95 (46,1%) 40 (46,0%) 55 (46,2%) 0,124

MetS risk factors (mean±SD)

Waist circumference (cm) 95,50 ± 12,56 97,62 ± 1,72 93,94 ± 12,97 3,8x10-2

DBP (mmHg) 80,67 ± 9,96 80,89 ± 10,02 80,52 ± 9,95 0,793

SBP (mmHg) 131,72 ± 20,02 133,00 ± 16,38 130,79 ± 22,33 0,245

HDL (mg/dL) 53,51 ± 13,33 49,61 ± 12,85 56,35 ± 12,99 2,5x10-4

TG (mg/dL) 107,71 ± 60,29 115,26 ± 74,73 102,19 ± 46,61 0,717

Glucose (mg/dL) 103,29 ± 33,91 109,79 ± 47,10 98,54 ± 1,66 2,6x10-4

MetS related diseases2

Hypertension 54 (26,2%) 20 (23,0%) 34 (28,6%) 0,054

Type 2 Diabetes 15 (7,3%) 8 (9,2%) 7 (5,9%) 0,796

Hypercholesterelomia 26 (12,6%) 7(8,0%) 19 (16,0%) 0,019

TOTAL 95 (46,1%) 35(40,2%) 60 (50,4%)

Medication

Hypertension 52 (25,2%) 19 (21,8%) 33 (27,7%) 0,052

Type 2 Diabetes 13 (6,3%) 6 (6,9%) 7 (5,9%) 0,782

Hypercholesterelomia 24 (11,7%) 6 (6,9%) 18 (15,1%) 0,014

TOTAL 89 (43,2%) 31 (35,6%) 58 (48,7%)

Smoking status

Current smokers 37 (18,0%) 19 (21,8%) 18 (15,2%) 0,869

Former smokers 42 (20,5%) 32 (36,8%) 10 (8,5%) 0,001

Never smokers 126 (61,5%) 36 (41,4%) 90 (76,3) 1,5x10-6

Regular Physical activity 80 (39,6%) 35 (40,2%) 45 (37,8%) 0,264

Table1. Characteristics of men and women participants (Data are presented as mean±SD for continuous variables and n (%) for proportions).

Genotype database

ID SNP1 SNP2 … SNP105

1 AA TG … CC

2 AT GG CA

3 TT TT … AA

… … … … …

… … … … …

… … … … …

208 TT GG … CC

105 SNPs Genotyped SNPs

INSEF sample:208

≈61000 Genotypes

Selected from

bibliography

Sequenom Genotyped

RFLPs Genotyped

Failed Successfull

Number of SNPs 114 73 41 9 105

0

20

40

60

80

100

120

10

1. Waist circumference Men≥94 cm Women≥80 cm

2. Blood pressure ≥130/85mmHg or Medication

3. TG ≥150mg/dL or Medication

5. HDL Men≤40mg/dL or Medication Women≤50mg/dL

4. Glucose ≥100mg/dL or Medication

Alberti et al, 2009

Metabolic Syndrome (MetS)

≥3 risk factors: MetS Diagnosis

Figure 1. Prevalence of MetS and its risk factors. Participants medicated for hypertension, hypercholesterelomia and diabetes were also accounted. Error bars represent the 95% confidence intervals. Abbreviations: MetS, metabolic syndrome; DBP, diastolic blood pressure; SBP, systolic blood pressure; HDL, high density lipoprotein cholesterol; TG, triglycerides

• Simple clinical tool for predicting diabetes and CVD and the conceptual basis for understanding at least part of the pathopgysiological link between metabolic risk, future diabetes and CVD;

• Provides a framework for reserach exploring a possible unifying pathophysiological basis for the observed cluster of risk factors;

• It can guide relative risk prediction and clinical management decisions;

• It provides na easily comprehensive public health message and reminds health professionals of the need to assess related risk factors when one risk factor is detected.

Why Metabolic Syndrome (MetS) approach?

Dichotomized MetS definition vs

Continuous MetS Score

Dichotomized definition of MetS enabling a yes or no diagnosis remains useful to clinical practice

BUT

in genetic epidemiological approaches , it reduces the statistical power of the MetS association studies and a

continuous MetS score will be a more appropriate alternative

Principal Components Analysis (Wijindaele et al. 2006)

6 Risk factors Wais circumference

Diastolic Bloood pressure Systolic Blood pressure

Glucose TG

HDL

MetS Score

Normalization

log10[log10(Systolic BP)]

1/[log10(Glucose)]10

[ln(HDL)]2

log10[log10(TG)]

Diferences between Sexes (T-test or Mann-Whitney test )

Normality Test (Shapiro-Wilk Test)

Higher MetS score =

Less favorable MetS profile

MetS Score

15

Continuous MetS score calculation by PCA analysis

Men

• PC1 and PC2 explain 35,9% and 27,4% of MetS score variance • measured correlations PC1 [PC2]: Waist circumference 0,650[0,255] Systolic blood pressure 0,826[0,057] Diastolic blood pressure 0,771[0,320] Glucose -0,598[0,147] HDL 0,079[-0,885] Triglycerides 0,305[0,818].

Women

• PC1 and PC2 explain 36,7% and 25,1% of MetS score variance • measured correlations PC1[PC2]: waist circumference 0,491[0,381] Systolic blood pressure 0,891[0,019] diastolic blood pressure 0,812[0,018] glucose -0,661 [-0,251] HDL 0,047 [-0,838] triglycerides 0,266 [0,770]

The result MS score was 0,00±1,41 in both genders

16

5. Results

Figure 2- Variation of the MSscore descriptive statistics according to the number of risk factors. The 5 risk factors considered are those presented

in the consensus MetS definition

MetS score increases progressively with increasing numbers of risk factors (ANOVA test p

17

Individual association with MetSscore (T-test)

Genotype Database

Candidate gene analysis

13 SNPs Type 2 Diabetes

7 SNPs Cardiovascular Diseases

9 SNPs Obesity

3 SNPs Dyslipidemias

5 Drug/Lipid metabolism

Multiple testing correction (Bonferroni Test)

18

5. Results

Table 2-List of SNPs selected in the present study. (Abreviations: MAF,Minor allele frequency

Gene NCBI ID Alteration Related traits European MAF1 Obtained MAF

CDKAL1 rs7754840 C→G Type 2 Diabetes 0,336 0,286

CDKN2A/B rs10811661 C→T 0,199 0,201

HHEX rs1111875 A→G 0,416 0,371

IGF2BP2 rs4402960 G→T 0,280 0,272

IL6 rs1800795 C→G 0,465 0,337

KCNJ11 rs5219 C→T - 0,333

KCNQ1 rs2237892 C→T 0,075 0,051

MTNR1B rs10830963 C→G 0,300 0,223

PPARG rs1801282 C→G 0,076 0,093

SLC30A8 rs13266634 C→T 0,239 0,286

TCF7L2 rs7903146 C→T 0,279 0,302

ADCY5 rs11708067 A→G 0,226 0,199

KCNQ1 rs231362 C→T 0,482 0,234

ACE rs4646994 Ins/Del Cardiovascular diseases - 0,420

NOS1AP rs12143842 C→T 0,188 0,265

ADRB1 rs1801252 A→G - 0,108

ADRB2 rs1042714 C→G 0,467 0,407

ADRB2 rs1042713 A→G 0,358 0,362

NOS3 rs1799983 G→T 0,342 0,417

NOS3 rs2070744 C→T - 0,451

19

5. Results

Table 2-List of SNPs selected in the present study (continuation).

Gene NCBI ID Alteration Related traits European MAF1 Obtained MAF

GNPDA2 rs10938397 A→G Obesity 0,446 0,481

MTCH2 rs10838738 A→G 0,363 0,282

NPC1 rs1805081 A→G 0,467 0,288

PTER rs10508503 C→T 0,092 0,075

SH2B1 rs7498665 A→G 0,382 0,303

FTO rs9939609 A→T 0,449 0,361

ADRB3 rs4994 C→T 0,088 0,090

GABRA2 rs279871 A→G - 0,434

TMEM18 rs6548238 C→T 0,150 0,127

APOE rs7412 C→T Dyslipidemia - 0,027

LDLR rs2228671 C→T 0,106 0,124

NPY rs16147 A→G 0,491 0,450

CYP2C8 rs10509681 C→T Drug/Lipid metabolism 0,137 0,129

CYP2C9 rs1799853 C→T 0,104 0,138

CYP2D6 rs16947 A→G - 0,393

CYP2C19 rs4244285 G→A 0,155 0,129

TPMT rs1142345 A→G 0,027 0,032

20

5. Results

Gene SNP ID Genotype n MetSscore P-value* Corrected P-value*

CYP2C19 rs4244285 GG 156 0,19±1,37 0,00044 0,016

AA+GA 50 -0,6±1,36

GABRA2 rs279871 AA 63 0,37±1,35 0,013 0,487

GG+GA 143 -0,16±1,41

NPY rs16147 AA 58 0,38±1,63 0,017 0,612

GG+GA 148 -0,15±1,29

TPMT rs1142345 AA 192 -0,07±1,38 0,0098 0,360

GA 13 0,97±1,63

Table 3. Polymorphism significantly associated with MetS score. (MetS score are presented as mean±SD).

*P-value were obtanied by T-test and Corrected P-value were obtained by Bonferroni Correction.

• Multiple linear regression Models

1 Age 2 Age+CYP2C19 3 Age+CYP2C19+GABRA2 4 Age+CYP2C19+GABRA2+NPY 5 Age+CYP2C19+GABRA2+NPY+TPMT ANOVA test p

Additive genetic effects

No association was found with environmental factors – lack of statistical power

Underlying phenotypes

GABRA2 rs279871 P=0.026 NPC1 rs1805081 P=0.039 NPY rs16147 P=0.024

GABRA2 rs279871 P=0.014 ADRB2 rs1042713 P=0.027 ADRB3 rs4994 P=0.040 CYP2C19 rs4244285 P=0.011

CYP2C19 rs4244285 P=0.014 ADRB2 rs1042713 P=0.027 ADRB3 rs4994 P=0.040

CYP2C8 rs10509681 P=0.017 MTNR1B rs10830963 P=0.048 ADRB3 rs4994 P=0.040 NPY rs16147 P=0.005

Multiple regression models only explain 5-10% of the phenotype variance

Additive genetic effects

ADCY rs11708067 P=0.0056 CYP2C19 rs4244285 P=0.0046 GABRA rs279871 P=0.0455 TMEM18 rs6548238 P=0.0078

CYP2C19 rs4244285 P=0.0366 TPMT P=0.0202

CYP2C19 rs4244285 P=0.0280 CYP2C8 rs10509681 P=0.0400 CYP2C9 rs1799853 P=0.0130 GABRA rs279871 P=0.0200 NPC1 rs1805081 P=0.0240 NPY rs16147 P=0.0350 TPMT P=0.0240

Multiple regression models only explain 5-10% of the phenotype variance

Conclusions • The quatitative MetS score has more power to detect association than the

tradicional MetS dichotomous definition;

• We have found a significant association between genetic variants in the CYP2C19, GABRA2, NPY and TPMT genes and the MetS quantitave score;

• Age + 4 genetic variants explain 23% of the MetS score variation;

• These genes are possibly involved in a pathophysiological mechanism responsible for the clustering of metabolic risk factors;

• No association is found between the phenotype using the traditional MetS definition and the analysed genetic variants;

• No association was found with environmental factors, likely due to lack of statistical power.

Genetic susceptibility to Influenza infection

• Infectious disease mortality risk has a heritable component. Children of

parents who died of an infectious disease are 6x more likely to die from an

infectious cause compared with the general population;

• An investigation of the influenza death records over the past 100 years in

the population of Utah provided evidence for an increased risk in close and

distant related relatives

• In some recent familial clusters of H5N1 infection, fatal cases curiously

clustered among relatives .

• More recently, a pilot study of host genetic variants associated with

influenza-related deaths among children and young adults has revealed that

individuals who died of influenza had low producing Mannose-binding lectin

2 (MBL2) genotypes conferring increased risk for Methicillin-resistant

Staphylococcus aureus (MRSA) co-infection .

Olsen et al., 2005

Ferdinands et al., 2011

Horby et al., 2010

Objectives

• To identify and characterize host and virus

genetic factors that influence susceptibility,

severity and outcome of 2009 pandemic

influenza A (H1N1) and to identify host-

virus additive and non-additive interactions

that would lead to increased susceptibility,

severity or outcome of this infection.

Materials and methods

1. Perform a case-control genetic association study using the nasal swab

samples that have been collected and sent to the INSARJ for diagnostic

purposes during the A (H1N1) pandemic on the context of the National

Influenza Surveillance Program (NISP) – targeted sequencing of

candidate genes;

2. Dissect viral genetic diversity by sequencing genomic segments of 150

virus present in randomly selected samples in each of the previously

established groups of influenza cases (mild and severe);

3. Analyze how host and viral genetic variation interact to influence disease

susceptibility and/or severity.

Study design

2009 Pandemics

Severe cases

(Hospitalized)

Mild cases

(Non hospitalized)

ILI H1N1 influenza virus

positive

96 (56M+40F) 212(115M+97F)

ILI - H1N1 influenza

virus negative

198 (110M+88F) 403 (217M+186F)

IFITM3 and influenza infection

Everitt et al., 2012

Everitt et al., 2012

INSA

• Vânia Gaio

• Vânia Francisco

• Ana Paula Gil

• Carlos Dias

• Mafalda Bourbon

• Astrid Vicente

• Baltazar Nunes

• Pedro Pechirra

• Raquel Guiomar

ARS Algarve

• Álvaro Beleza

• Francisco Mendonça

• Filomena Horta Correia

• Aida Fernandes

Acknowledgements

PTDC/SAU-ESA/101743/2008